study of advance directives in SUPPORT, Teno and colleagues (1998) found that advance directives were invoked and often played a role in decision making. However, directives were invoked only when the patient was “hopelessly ill.”
Moving this decision upstream from a point when treatment is judged almost certainly futile will take a fundamental change in the dialogue that occurs between patients, families, and physicians. Discussion of prognosis at an earlier stage must be accomplished in such a way that it does not shatter hope, yet allows a dying person to make realistic choices about medical care. Cancer, unlike other leading causes of death, does have a relatively predictable disease trajectory that would allow for such dialogues to be developed and implemented.
The lack of impact of the SUPPORT intervention (which provided physicians with information on patient preferences and prognoses but did not result in increased physician understanding of patient preferences, timing of DNR orders, reduction in days spent in undesirable outcome states, and reduction in resource utilization) has been taken by some as a rationale for endorsing “glide paths” (i.e., “default pathways”), rather than finding better ways to communicate and, ultimately, implement patient self-determination. A careful review of the SUPPORT findings, however, suggests that the intervention itself was inadequate to improve communication, not that improved communication is impossible.
Given the existing research, the only firm conclusion that can be drawn is that communication is lacking between physicians and patients (Haidet et al., 1998; SUPPORT Principal Investigators, 1996) and that physicians often misunderstand patient preferences (Teno et al., 1995). Research in communication and decisionmaking has focused largely on the last days of life. The more sentinel decisions, though, may be stopping active treatment or choosing palliative chemotherapy, radiation therapy, or surgical treatment earlier in the course of illness. The evidence base to support guidelines for these decisions is preliminary at best. Research on how best to communicate this information to patients has only begun.
NCI and AHRQ could fund research to develop the evidence base for palliative chemotherapy, radiation, and other treatment modalities. Such research should consider the treatment’s meaning to patients and families, toxicity, and impact on quality of life.
Cooperative Oncology Groups could standardize measures and schedules to examine both treatment toxicity and quality of life. This would facilitate meta-analyses to develop the evidence base for palliative chemotherapy, radiation, and surgical treatments.